News

Invitation to Online Course – "Cybersecurity Forensics Series I: Penetration, CTF Strategy, and OSINT Practical Reconnaissance" on June 5, 2026.

OpenClaw (commonly known as "Lobster"), a recently widely acclaimed open-source AI Agent software tool, functions as an autonomous artificial intelligence virtual assistant capable of executing tasks on behalf of users. It is primarily deployed on enterprise local servers or PCs and has the capability to call other Large Language Models (LLMs) and Application Programming Interfaces (APIs). By receiving text commands from users via instant messaging platforms, it can perform tasks such as scheduling, sending messages, organizing files, and writing code. Through OpenClaw, users can create their own personalized AI Agent tools (commonly referred to as "raising lobsters").

However, because these tools possess extremely high system privileges and operate autonomously 24/7, they can easily become a major breach for hackers to infiltrate personal hosts and corporate networks if proper defense mechanisms are not deployed. This could lead to the leakage of users' personal data and sensitive information—such as account credentials, passwords, and financial data—posing risks of identity theft and financial loss.

The Administration for Cyber Security (ACS) of the Ministry of Digital Affairs (MoDA) has issued a press release titled "Reminding Organizations to Implement Five Cybersecurity Defenses When Deploying OpenClaw" (hereinafter referred to as "the Press Release"). It reminds the business community of the specific cybersecurity threat scenarios they must be highly vigilant about, as well as the essential cybersecurity protective measures that must be implemented while evaluating and utilizing such open-source tools to enhance operational efficiency.

Please refer to the attachment for the Press Release, or visit the official MoDA ACS link: https://moda.gov.tw/ACS/press/news/press/19294

The IT Committee (Information Security Committee) hereby forwards this press release to serve as a reference guide for group executives when evaluating AI agent tools and deploying related cybersecurity defenses.

Should executives from group affiliates have any questions or suggestions regarding topics related to the revised clauses of the cybersecurity guidelines, please feel free to contact the IT Committee event coordinators:

• Duke Huang: (02) 7727-2650
• Eva Liao: (02) 7727-2352

[Activity Brochure 1]